Looping in R to perform a GSEA analysis

Question

[rodolfo.peacewalker]

Hi!

I need to run a GSEA analysis on distinct lists of pre-ranked DEG, as well as diverse lists of gene sets obtained from MSigDB. To run the analysis using one pre-ranked list and one gene set, the code is:

fgsea_results <- fgsea(kegg.db, #Gene set
                                   b, #List of pre-ranked DEG
                                   minSize = 15, 
                                   maxSize = 500, 
                                   nPerm = 10000
                                   )

To follow good practices in code writing and, also, for avoid DRY I figured out that I need to build a for loop to run the analysis in all my pre-ranked lists or the gene sets. A first attempt to build a for loop was the next

for (samples in c(a, b, c, d)) {
   fgsea(kegg.db, #Gene set
             samples, #List of pre-ranked DEG
             minSize = 15, 
             maxSize = 500, 
             nPerm = 10000
              )}

However, there's something I missed cause result was not successful. I need your help to solve this issue.

Best regards! Rodolfo

[hafsaiiuc39]

fgsea_results <- fgsea(kegg.db, #Gene set b, #List of pre-ranked DEG minSize = 15, maxSize = 500, nPerm = 10000 )

Here, would you tell me please how to create kegg.db? I mean how to retrieve pathways for given differentially expressed gene?

[ATpoint]

Please make comments with ADD REPLY. You can download KEGG pathways e.g. from MSigDB as gmt files and then read these into R with fgsea:: gmtPathways().

Answer 1

[ATpoint]

If you c() lists then you concat them together. By the way c is an unfortunate variable name because there is a function called c().

Try:

for (samples in c("a", "b", "c", "d")) {
   fgsea(kegg.db,
         get(samples), 
         minSize = 15, 
         maxSize = 500, 
         nPerm = 10000)
}

[rodolfo.peacewalker]

Thanks for your help and suggestions! It worked fine.